The first time I read WAR next to a player’s name, with the implication that it was supposed to tell me something, in my head I pronounced it like “star.” In my mind’s eye I can see myself slowly mouthing W…A…R. It felt strange to think about someone’s “war,” which I suppose is pronounced like it is in “world war.” It didn’t help that I had no idea what it meant and had even less of an idea of what to do with the number. Learning that it stood for “Wins Above Replacement” didn’t initially help. If you count yourself among that group, I’ll summarize WAR in a moment. For now, it’s important to remember that WAR is not just a product of what has been termed the “sabermetric revolution” in baseball, but that it is (right now, at least) the apogee of it. But thinking in terms of WAR is not limited to the initiated cadre of “sabermetricians” or consumers of their baseball thinking. Rather, WAR provides a dynamic new way of doing the only thing about baseball that truly is timeless: telling stories. WAR invites a macro perspective of individual performances over the course of seasons and careers that helps us narrate baseball in a compelling manner. The link between statistics and storytelling illustrates that any division that might exist between “old school” and “new school” ways of observing and consuming baseball is artificial. Indeed, they reinforce one another.
First things first: what is WAR? As FanGraphs’ Dave Cameron posits, WAR is a metric that attempts to answer a simple question that everyone has asked: “how good is that player?” Every statistic answers a slice of that question, but WAR’s purpose is to tie every component of a player’s game together into a single number. The building blocks of WAR for position players are a variety of metrics that measure the ability to create runs with the bat (which I explain here), prevent runs on defense (explained here, by me), and navigate the basebaths (cogently explained by another person here). Fielding Independent Pitching (explained, by me, here) is the essential component for pitchers. Notably, WAR is adjusted to account for defensive position. This means that Carlos Gonzalez, center fielder, would have a different WAR from Carlos Gonzalez, left fielder, because center is more difficult to play.
But what does it look like? The other question that the statistic answers helps us visualize WAR. It asks, “how much more value does that player provide compared to the dime a dozen AAA player that would replace him in the event of a comical off-field injury?” Replacement level is zero, and anything above is calculated value above your everyday Triple-A warm body. In 2013, the position player who received the most playing time and ended the season with a WAR of zero was Yorvit Torrealba. Wilin Rosario, on the other hand, ended the season with a WAR of 2.2, which roughly means that he was a solid major leaguer. Rosario created runs seven percent better than league average, but we have to recall that defense plays a role as well. Wilin Rosario is, right now, not a very good defensive catcher, and that hurts his overall value. Still, the Rockies would have lost the value of about two wins if Rosario sprained his ankle because he fell off of an overturned bucket while changing a light-bulb, forcing Torrealba into the daily line-up. It’s also possible to have negative WAR—having negative value is obviously not good. The Rockies had several negative WAR players in 2013, and the lowest was the team’s third catcher, Jordan Pacheco, who clocked in at -1.4 WAR. If an injury befell Torrealba—for instance, if Rosario fell on Torrealba while falling off of a bucket while changing a light-bulb—the Rockies would have been in rough shape at catcher, as Pacheco offered the team negative value given his poor hitting, below average defense, and the necessity to play a difficult fielding position. It would have been a sad story, indeed.
WAR is a good story statistic because it accumulates over time, but not necessarily in a linear fashion. To elaborate upon this, I’ll look at Troy Tulowitzki. I mentioned that Rosario’s 2.2 WAR was about league average, but we also have to keep in mind catchers don’t play every day, so he accumulated his WAR in 121 games and 466 plate appearances. But that doesn’t necessarily elevate Rosario’s value too much. In 2013, Troy Tulowitzki reached lofty heights despite limited playing time. He had 5.6 WAR in 126 games and 512 plate appearances—anything in excess of five WAR is illustrious, and beyond that flirts with MVP candidacy. Tulowitzki’s 2013 story fits into the story of his career. He was a star right away, compiling 5.2 WAR in his rookie season in 2007. In 2007 he also played the most games he has ever had in a single season, 155, and accumulated the most plate appearances, 682. So far, 2007 was his fifth best offensive year. Tulowitzki produced runs nine percent better than league average, roughly in line with Rosario’s 2013. Good, but not spectacular. Tulowitzki’s defense, however, was superlative. He saved 31 runs in 2007, which was 13 runs ahead of second place among major league shortstops.
After 2007, Tulowitzki’s offense has been ascending, while his defense has been in decline. Defense tends to peak earlier than hitting, so this is unsurprising. What has remained consistent, however, is Tulowitzki’s value. WAR tells us that his progressive development as a hitter compensates for his regressing defensive play. Tulowitzki’s best season according to WAR was 2010, with a mark of 5.9. He produced runs 40 percent better than league average (compare that to nine percent in his rookie season). Tulowitzki played very good defense that year as well, saving 19 runs. Even more notable, he did this in just 122 games, compared to 155 in 2007. The story of his offense and defense persists. In 2013, in 126 games, Tulowtizki saved six runs (still quite good), while with the bat he created runs 43 percent better than league average.
Even the WAR in Tulowitzki’s lost seasons, 2008 and 2012, tell the story of a baseball player who was changing. Two-thousand-eight was, by a fair margin, Tulowitzki’s worst season. Offensively, he produced runs 17 percent below the average major leaguer. He was on the right side of the baseline defensively, but his value there was limited to just three runs. The result was a slightly better than replacement performance: 0.7 WAR. Most notably, he did this in 101 games. The only season in which Tulowitzki did not play 100 games was 2012, when he only played 47. It was also the only season in which he was a negative defender, costing the Rockies six runs. But he still hit, just not as well as his healthy self, as Tulowitzki created runs at just fifteen percent better than league average. And yet, his 2012 WAR was nearly double his 2008 WAR: 1.3 compared to 0.7. He produced more WAR in less time while providing slightly below average defense at the most difficult spot on the field.
So what does Tulo’s WAR story tell us about his future? His defensive value is never going to return to 2007 form, but that’s alright, because his value is shifting more and more to his bat. It’s still quite possible that he has yet to peak offensively. But one of the more important things WAR tells us is that his career is not one of ascendance followed by decline. He has moving parts within his own skill-set, and if one of his skills sees dramatic movement rather than the slow migration that has taken place thus far, his total value will also change. Going forward, I would look for Nolan Arenado to replicate Tulowitzki’s career arc. Arenado probably won’t have a defensive season as good as his 2013 rookie year, but his bat will get better, even if it’s never Tulowitzkian. Paying attention to his WAR and the pieces that contribute to making it will, one day, help us narrate his career as well.
Of all of the new statistics used in baseball, WAR is the most polemical. Last year, ESPN’s Jim Caple wrote an impassioned plea to use WAR with caution. It should not, he warned, be an indisputable trump card. He is right. WAR is useful, but no panacea. More recently, upon Derek Jeter’s retirement, Allan Barra much less delicately offered the following imperative: “Shut up, stat nerds and haters: You’re wrong about Derek Jeter.” While not directed at WAR specifically—but instead at statistically inclined and pocket protector adorned observers—Barra’s tone is an example of the artificial division that exists between baseball fans who prefer and trust different metrics. I like WAR because it condenses a player’s total value to an easily digestible number. Not only that, and as Jeff Sullivan points out, it also contextualizes the value individual players have for their team. Just as the component parts of a player’s game contributes to his value as determined by WAR, each individual player is, in the end, a small contributor to a team’s success. I also like WAR for the same reason I like Runs Batted In (RBI): they both contribute to telling stories about baseball. WAR does it on a macro scale, while RBI does it on a micro scale.
There are a variety of ways to observe and consume baseball. Creating artificial barriers among them halts conversation, to the benefit of nobody. We are all speaking the same language—there’s just a bit of idiomatic variation, as well as pronunciation.